Closed-Loop Control Modalities in Type 1 Diabetes: Efficacy and System Acceptance In 2009 we initiated one of the first NIH studies dedicated to engineering and clinical testing of closed-loop control (CLC) of type 1 diabetes. Since then, we have achieved key milestones and derived conclusions which enabled further research in this rapidly growing field. Notably, we proposed the idea that the artificial pancreas is not a single all-in-one device but a network encompassing the patient in a digital treatment ecosystem that can offer and alter different treatment modalities in real time depending on the patient's clinical state. This new notion was reflected in: (1) Our modular engineering design of CLC algorithms, which now allows various treatment modalities to be initiated and swapped without interruption; (2) The Diabetes Assistant (DiAs) - the first portable CLC hub using a smart phone to run control algorithms and specifically designed to be operated by the patient, which is now used in a number of outpatient studies in the U.S. and in Europe, and (3) The Unified Safety System (USS Virginia) - the first CLC algorithm engineered to adapt its mode of operation during the course of every night, first mitigating after-dinner hyperglycemia and then sliding the patient to a target morning glucose of 120mg/dl, thereby resetting his/her metabolic state for a new day. Using these technologies, we now propose to compare in a randomized cross-over trial the long-term efficacy of three treatment modalities - sensor-augmented pump (SAP) vs. USS+SAP during the day(d) vs. USS+CLC(d). We plan to randomize 84 patients with type 1 diabetes into two different treatment sequences: SAP?USS+SAP(d)?USS+CLC(d)?USS+SAP(d) and USS+SAP(d)?USS+CLC(d)?USS+SAP(d)?SAP. Each treatment modality will continue for 2 months - sufficient time to address the following specific aims: SA1: Overnight CLC achieved by USS+SAP(d) will be superior to SAP alone in terms of: (1) Improved HbA1c without increasing the risk for hypoglycemia; (2) Reduced incidence and risk for hypoglycemia overnight, and (3) Reduced fear of hypoglycemia and improved diabetes quality of life scores. SA2: CLC during the day achieved by USS+CLC(d) will preserve the benefits of USS+SAP(d) and will be superior to USS+SAP(d) in terms of: (1) Increased time within target range of 70-180mg/dl during the day; (2) Reduced risk for hypoglycemia during and after exercise, and (3) Reduced postprandial glucose variability. SA3: CLC system acceptance evaluated by focus-group interviews and technology acceptance scores will be: (1) Superior, for USS+SAP(d) compared to SAP alone, i.e. adding USS overnight will increase patients' acceptance of CLC, and (2) Marginally inferior, for USS+CLC(d) compared to USS+SAP(d); i.e. some patients would prefer SAP alone during the day due to perceived increased system complexity. Overall, we expect to establish that a distinct overnight CLC modality (USS Virginia) combined with SAP therapy during the day is a viable precursor to future adaptable therapeutic schemes, achieving glycemic control that is superior to SAP alone and optimal balance between system complexity and perceived benefits.